Sand

Since the 1970s, the integration of nobake binders into green sand systems has been investigated significantly. By updating a previous conclusion, closure to the dilemma might be at hand. Rodney L. Naro, ASI International Ltd., Cleveland

Binding Past Knowledge Amid further studies of phenolic urethane binders’ performance, some of the investigations have contested one another. Despite this closer examination and a dichotomy of information, one important question remained: “What effect will core butts and shakeout core sands have on green sand properties as they enter a green sand molding system?” Several investigators have looked into possible chemical-related effects of core sand contamination on green sand properties. One study concluded that although specific change to a green sand system might be slight—and in some cases even advantageous—the long-term effects might be grave. In those cases where an effect was noticed, it was felt that condensed resin distillates (a byproduct of binder pyrolysis) impeded the bonding effectiveness of bentonite. The effects observed were most apparent in the deterioration of both green compression and wet tensile Table 1. Moisture Content of New/Recycled Sand Blends strengths of the molding sands. % Contamination Other investigations have Binder System 1% 5% 10% 25% 50% 100% Average Std. Dev. declared the benefits of mixFNB-Phos 2.70 2.72 2.82 2.74 2.84 2.72 2.76% 0.05% ing, such as little-to-no differFNB-TSA 2.94 2.86 2.88 2.86 2.88 2.88 2.88% 0.03% ence between the rebonding FNB-BSA 3.00 2.96 2.74 2.82 3.00 2.80 2.89% 0.10% characteristics of 100% rePhenolic Nobake 2.70 2.64 2.76 2.94 3.00 2.85 2.82% 0.13% cycled phenolic urethane Phenolic Urethane 2.96 2.96 2.96 2.82 2.84 2.70 2.87% 0.10% coldbox (PUCB) process sand Silicate Nobake 3.00 2.84 3.00 2.98 3.00 2.86 2.95% 0.07% and new sand. A similar study Standard Green Sand System Moisture Content: 2.87% ± 0.11% revealed that 15% particulated core sand from both phenolic Table 2. Available Clay Content of New/Recycled Sand Blends (Determined from IMC Charts) hot box and PUCB cores could % Contamination be blended into a green sand Binder System 1% 5% 10% 25% 50% 100% Average Std. Dev. system with minimal effect on FNB-Phos 5.50 5.60 5.60 5.30 6.10 5.50 5.60% 0.24% molding properties. The lossFNB-TSA 5.60 5.30 5.70 5.80 5.70 5.80 5.65% 0.17% on-ignition (LOI) values of the FNB-BSA 6.00 6.10 5.60 5.70 6.00 6.20 5.93% 0.21% particulated sands were 2.13%. Phenolic Nobake 5.80 5.50 5.70 6.10 6.10 6.00 5.87% 0.22% A different investigation also Phenolic Urethane 5.70 6.00 6.00 5.80 5.70 5.50 5.78% 0.18% focused on the effects of core Silicate Nobake 6.00 6.00 6.20 6.10 6.00 6.50 6.13% 0.18% sand dilution with PUCB bindStandard Green Sand System Available Clay: 5.80% ± 0.20% ers. This study concluded that ince 1979, the growth of chemical binders—particularly phenolic urethane binders—has been phenomenal. Although usage of most chemical binders has increased during the last 25 years, the use of phenolic urethane binders has grown at the fastest rate. In 1971 the U.S. metalcasting industry used only 2.7 million lbs. of coldbox and nobake phenolic urethanes. In 2003, 150 million lbs. of both resins were projected to have been consumed in the U.S. Estimated worldwide use is considered to be greater than 300 million lbs. These newer binder systems have helped to meet the increased demands of the metalcasting industry for a variety of reasons including improved dimensional accuracy, increased productivity and reduced energy consumption. However, spanning the same time period has been an ongoing controversy about the effect of recycled chemi-

34

cally bonded sands on green sand properties. In the 1970s, few reports were conducted on the topic. But over the last 25 years, more engineers have investigated the issue, mainly focusing on phenolic urethane binders, and in particular, whether their introduction into a green sand system affects rebonding properties. One of the earlier studies (conducted in 1979) investigated the effects of chemical binder core sand contamination on the properties of a bentonitebonded green sand. Although the results were relevant at that time, the study was updated this year, reviewing the same tests with additional parameters. Also, the updated investigation studied three different metalcasting facilities and how phenolic urethane binders affect their green sand systems. This article examines both the 1979 and updated studies of such binders as well as results of other investigations found during this 25-year span.

MODERN CASTING / August 2004

Table 3. Bonding Clay Content of New/Recycled Sand Blends

Then and Now By updating the research from 100% Average Std. Dev. 25 years ago, a more detailed 3.30 3.47% 0.26% approach can be taken in re3.60 3.38% 0.28% gards to this controversial topic. 4.40 3.85% 0.34% The 1979 study dealt exclusively 4.10 3.88% 0.13% with residual chemical effects, 4.10 3.93% 0.15% such as acidity and basicity con3.90 3.85% 0.15% tributions from recovered shakeout sand. It did not examine physical interactions, such Table 4. Available Clay Content of New/Recycled Sand Blends (Determined by Methylene Blue Titration) as pyrolysis condensate residues % Contamination on sand grain surfaces and/or Binder System 1% 5% 10% 25% 50% 100% Average Std. Dev. condensate residues on bentoFNB-Phos 7.60 7.80 8.00 7.60 8.00 7.60 7.80 0.18 nite particles from binder deFNB-TSA 7.80 7.80 7.80 7.80 8.00 7.80 7.84 0.08 composition. That study showed that recycled nobake core sand FNB-BSA 8.00 7.60 8.00 7.80 7.80 7.80 7.80 0.13 additions up to a 50% substituPhenolic Nobake 7.60 7.60 7.40 7.80 7.60 8.00 7.68 0.20 tion rate and after exposure to Phenolic Urethane 7.80 7.80 7.80 7.60 7.80 7.60 7.72 0.10 casting temperatures had littleSilicate Nobake 8.20 8.20 8.60 8.60 8.40 8.60 8.48 0.16 to-no effect on green sand propStandard Green Sand System Available (MB) Clay: 7.8% ± 0.20 erties. In all cases, optimum green sand properties were note high-percentage LOI sands had the spent PUCB sands, regardless of achieved after 25 min. of mulling. such a coating. LOI value, had no effect on green With a few exceptions, green sand However, lustrous carbon doesn’t properties of the molding sand. properties, such as compactability and burn in a LOI test, hence, lustrous Still, examinations have been pergreen compressive strength, for the carbon and LOI values are not related. formed announcing that mixing core bentonite-bonded coldbox recycled Therefore, the intermediate LOI values and green sands will have significant shakeout sand were essentially equivawere likely the result of some other effects on molding properties. Because lent to those of the new base sand. pyrolysis condensate residues. If these the curing mechanisms of all chemical Although some slight shifts in perforcondensates from intermediate therbinders involve various modes of acidmance were observed, these deviamal decomposition levels produce a base catalysis, concern exists on how tions were usually within one standard slippery or oily film on the sand grains, residual pH changes from shakeout deviation of the average base green then one would expect to see reduced core sands affect clay-bonded sand sand system properties. ability for bentonites to bond to sand properties. Past studies have revealed In the current study, the results from surfaces. This may be partially overthat pH significantly affects green sand 1979 were reviewed to determine the come by additional mulling. properties. These studies note that as effect of deliberate additions of reThe pH and lustrous carbon dilemthe pH value of green sand increased, cycled shakeout core sands and the mas will be discussed later in this article. green compression strength decreased effects on the properties of the new In addition to chemical contaminawhile dry strength, permeability and bentonite-bonded molding sand. To tion, another factor affecting green sand flowability increased. investigate these effects, varying properties is the physical effect from Reconditioning of green sand sysamounts of recycled nobake sands were additional core sand entering a green tems has been said to be one of the added to fresh sand mixtures to detersand system. These sands may be more most difficult steps in molding-sand mine possible interactions between the difficult to mull with bentonite if they technology. Bentonite and water prefrecycled core sand and new sand syscontain pyrolysis condensate residues. If erentially settle on the molding sand tem. Green compressive strength, shear appropriate tests are not run to make up grains already coated with clay. Long strength, compactability and permefor deficiencies in bentonite content, and intensive mulling is required beability were measured in an attempt to loss of bonding properties could result. fore recycled core sands assume the determine potential chemical interacsame properties as those of the base molding sand. Table 5. Effect of Core Sand Contamination on Compactability One investigation found that % Contamination incorporating lustrous carbon Binder System 1% 5% 10% 25% 50% 100% Average Std. Dev. forming additives at high levels FNB-Phos 46.00 46.00 55.50 48.00 48.00 47.25 48.46 3.25 to improve the refractoriness of FNB-TSA 52.00 53.00 50.00 51.00 50.00 47.25 50.54 1.82 sands often leads to the buildup FNB-BSA 49.50 49.75 48.50 49.75 49.00 45.75 48.71 1.40 of an oily film, which further Phenolic Nobake 50.00 48.75 50.00 47.50 46.75 42.50 47.58 2.57 reduces bentonite swelling caPhenolic Urethane 53.00 52.75 51.00 50.75 49.50 47.00 50.67 2.03 pacity. Lustrous carbon, several Silicate Nobake 53.00 51.00 46.25 44.75 46.25 33.75 45.83 6.13 engineers have claimed, hinders Standard Green Sand System Compactability: 54% ± 4% green sand properties, and some % Contamination Binder System 1% 5% 10% 25% 50% FNB-Phos 3.20 3.60 3.60 3.20 3.90 FNB-TSA 3.10 3.30 3.00 3.80 3.50 FNB-BSA 3.80 3.90 3.50 3.40 4.10 Phenolic Nobake 3.90 3.70 3.80 4.00 3.80 Phenolic Urethane 3.90 4.00 4.10 3.80 3.70 Silicate Nobake 4.00 3.90 4.00 3.70 3.60 Standard Green Sand System Bonding Clay: 3.90% ± 0.10%

MODERN CASTING / August 2004

35

tions. Other properties measured Table 6. Effect of Core Sand Contamination on Green Compression Strength % Contamination were moisture, available clay, Binder System 1% 5% 10% 25% 50% 100% Average Std. Dev. mean available clay and bondFNB-Phos 19.00 19.30 17.50 16.90 21.00 17.35 18.51 1.42 ing clay contents. FNB-TSA 15.65 16.40 17.20 19.30 18.05 19.05 17.61 1.33 The procedure used was diFNB-BSA 18.95 20.90 18.00 17.70 18.87 24.20 19.77 2.23 vided into two phases. The first Phenolic Nobake 20.50 19.95 19.75 21.50 20.65 22.50 20.81 0.94 phase consisted of generating Phenolic Urethane 18.65 19.20 21.20 19.40 19.00 16.65 19.02 1.33 shakeout sand from chemically Silicate Nobake 19.80 20.13 21.95 19.93 19.50 24.85 21.03 1.88 bonded molds after performing casting trials. The second phase Standard Green Sand System Green Compression Strength: 19.0 psi ± 1.0 psi evaluated the effects of the recycled sands when added in varifact that facilities B and C use aromatic properties of recycled core sand and ous proportions to a new sand-clayand aliphatic solvents, which are nornew green sand. water mixture. mally associated with PUCB systems. When this investigation was moved The sand was a blend of new sand On the contrary, facility A used a to three metalcasting facilities (A, B and recycled core sands obtained during system based on bio-diesel solvents and C) the results were all similar. casting shakeout. Recycled core sand due to a “brittle-sand” condition that Facilities A and B, both of which use contaminant levels of 1%, 5%, 10%, 25%, resulted in casting surface deteriorarigid flasks, run nearly identical green 50% and 100% were evaluated. (The tion. After converting to the new syssand properties, whereas facility C, 1979 research did not include contamitem, the facility no longer needed to which utilizes vertically parted molds, nant levels of 1% and 100%). add new sand to the green sand system requires more robust green sand The Current Event to restore properties and, like facilities properties. Even though two entirely B and C, did not report any green sand different base sands were used at With the addition of the 1% and property deterioration. facilities A and B, the physical prop100% contaminant levels, one finding erties of the green sand system were was as noticeable in the current study Truth Be Told remarkably similar. as in 1979—almost all of the binder By evaluating the reports from the Facilities B and C reported that PUCB systems for the various properties studthree facilities as well as the updated shakeout sand re-entering their green ied were found to be within a standard laboratory research, these investigasand systems did not present any probdeviation of 1 (Tables 1-7). This proves tions advance the argument that lems. Both PUCB shakeout sands had that the properties of a contaminated nobake sands do not significantly relatively low LOI values and both system are similar to those of a stanaffect molding characteristics of a facilities treated spent PUCB shakeout dard base green sand system. The only green sand system. sand entering the sand system as a property where more than two binder pH level—Even though certain nobake new sand addition; they added the systems exceeded the 1 standard debinders may contain strong acids or appropriate amount of bond and waviation level was permeability, which bases in either the resin or the catalyst, it ter to compensate for the PUCB sands. had a standard deviation of 10 (Table is evident that after the binders have These findings are likely due to the 8). Tables 9 and 10 examine various undergone curing and casting, they impart very little of their Table 7. Effect of Core Sand Contamination on Green Shear Strength original acidic or basic character % Contamination to the reclaimed sand. Binder System 1% 5% 10% 25% 50% 100% Average Std. Dev. When such sands enter into a FNB-Phos 4.80 4.90 5.50 3.60 5.40 4.50 4.78 0.63 green sand molding system, they FNB-TSA 4.35 5.30 4.10 5.15 5.05 5.00 4.83 0.44 do not significantly reduce the FNB-BSA 4.70 5.27 4.57 5.37 5.27 5.90 5.18 0.44 bonding effectiveness or moldPhenolic Nobake 5.50 6.45 5.20 5.60 5.10 4.75 5.43 0.53 ing properties of bentonite clays in the mix, thus, there is no Phenolic Urethane 4.35 5.60 5.60 5.20 5.70 4.50 5.16 0.54 correlation between residual pH Silicate Nobake 5.75 5.75 5.65 5.45 4.93 5.00 5.42 0.34 and green properties. Standard Green Sand System Green Shear Strength: 5.10 psi ± 0.70 psi There are several reasons why residual binders on reclaimed Table 8. Effect of Core Sand Contamination on Permeability nobake sand do not impart ap% Contamination preciable acidic or basic impuBinder System 1% 5% 10% 25% 50% 100% Average Std. Dev. rity functionality to a green sand FNB-Phos 112.00 112.00 117.00 82.00 94.00 72.00 98.17 16.84 system. In some cases, during FNB-TSA 103.00 100.00 85.00 91.00 85.00 78.00 90.33 8.79 the curing reaction between resin FNB-BSA 91.00 91.00 88.00 83.00 94.00 75.00 87.00 6.35 and catalyst, acidic or basic subPhenolic Nobake 110.00 112.00 112.00 111.00 97.00 85.00 104.50 10.18 stances react to form neutral Phenolic Urethane 113.00 114.00 108.00 108.00 111.00 77.00 105.17 12.80 salts, which are relatively nonSilicate Nobake 102.00 105.00 98.00 107.00 109.00 79.00 100.00 10.03 reactive with bentonite clays. Standard Green Sand System Permeability Index: 107 ± 10.0 In addition, acidic or basic 36

MODERN CASTING / August 2004

• if possible, reducing catalysts in a nobake Table 9. Recycled Shakeout Core Sand Properties Core Sand System Contaminant pH ADV Loss on Ignition (%) core weight to provide system are normally Furan Nobake—Phosphoric acid 2.8 -15.2 0.81 higher levels of used in such small Furan Nobake—Toluene Sulfonic acid 4.3 +3.2 0.78 thermal breakdown. amounts that after castFuran Nobake—Benzene Sulfonic acid 5.2 +4.0 1.08 Other factors that ing and reclamation, Phenolic Nobake—Benzene Sulfonic acid 5.2 +5.2 1.03 metalcasters using only a minimal amount Phenolic Urethane Nobake (PUN) 7.0 +3.8 0.62 PUCB binders should of residual catalyst reSilicate Nobake (SNB) 8.5 +5.4 0.17* consider are that, in mains on the sand. It New Sand Base Reference 6.3 +3.0 0.06 the investigation, appears likely that in PUCB binders formuinstances where there *Conventional LOI determination is not applicable to inorganic systems. lated with aromatic have been reports of and aliphatic soldecreased green sand vents generally did not result in on the proposed mechanism of lusproperties because of suspected PUCB green sand property deterioration. trous carbon formation, lustrous carcontamination, pyrolytic condensate resiIn instances where the deteriorabon films from PUCB binders cannot dues on either sand grain surfaces or tion of green sand properties is be responsible for some reports of bentonite particles, from incomplete linked to the use of standard PUCB diminished green sand properties. If pyrolysis, may affect resultant green sand binders, the facility should conand when green sand properties deteproperties. These condensates may imsider coldbox binder systems that riorate from PUCB binders, it probably pede the swelling action of bentonite use bio-diesel solvents. results from unique thermal circumand, for a given mulling time, the smearFurther, appropriate additions of stances occurring within the core during action of bentonite particles needed new bentonite clay and water should ing casting, not lustrous carbon. Such to develop green sand properties. be made to adjust total clay back to problems may be overcome by emLustrous Carbon—Lustrous carbon its original value. Thermal conditions ploying longer mulling cycles. defects have been blamed as a cause of within the mold may result in incommany contamination problems. Some Useful Sand to the Core plete combustion of PUCB decominvestigators of core sand contaminaposition gases during pouring. ConDespite the 25 controversial years tion have concluded that lustrous cardensate residues generated from such regarding green sand contamination, bon is one cause of reported deterioraconditions may inhibit the ability of this recent investigation augments the tion green sand molding properties. bentonite particles to effectively theory that both recycled core and Lustrous carbon, a brittle material meacoat sand grains. green sands are compatible in the same suring 0.0001 in., forms at the moldLastly, past studies have claimed the system. Although this study showed metal interface in binder systems that addition of small amounts of sodium that reclaimed nobake core sand had contain high levels of carbon and relacarbonate or soda ash (Na2CO3) to a little acidic or basic interaction with tively low levels of oxygen. It cannot bentonite clays, such sand should be coat sand grain surfaces. bentonite green sand system has helped treated as new, unbonded sand. As additional metal flows into the improve the bonding action of bentoHowever, there are a number of mold, these films may become dislodged nite clays as well as soda ashes’ ability methods that may be used if green and be flushed ahead of the leading to cleanse sand grain surfaces. sand properties need to be restored. edge of the incoming metal stream. If Although green sand systems might These include: the films are not dissolved in the metal falter as a result of other conditions, • increasing mulling time; or oxidized, solidification can proceed after this thorough examination, no • increasingwestern bentonite levels; against these accumulations, resulting in indication was found that recycled • increasing mold venting to help release surface wrinkling characteristics of luschemically bonded sands have any gaseous decomposition products; trous carbon defects. deleterious effect on the properties of • reducing binder content in cores; However, the lustrous carbon-forma green sand system. MC • scalping or removing core butts from ing tendencies of certain chemical bindThis article was adapted from paper the shakeout system; ers are harmful only if large amounts of (04-001) presented at the 2004 Metalcasting Congress carbon films form and then are dislodged from the mold- Table 10. Standard New Green Sand Properties About the Author metal interface during pourProperty Average Value Standard Deviation Dr. Rodney L. Naro is the president ing. When this happens, wrinand CEO of ASI International Ltd., Moisture 2.87% 0.11 Cleveland. He recently received the kling and surface laps result. Bonding Clay 1 3.9% 0.10 AFS Joseph S. Seaman Gold Medal If the lustrous carbon films for technical contributions to the Available Clay 1 5.8% 0.20 metalcasting industry. are not dislodged during fillAvailable (MB) Clay 2 39 ml 1 ing of the mold cavity, lusFor More Information Green Shear 5.1 psi 0.7 trous carbon formation may “Evaluating Refractory Coatings for Green Compressive 19 psi 1 PUCB Binders: A Practical Approach,” actually improve casting surPermeability 107 10 S.G. Baker, 2002 AFS Transactions face finish. paper No. 02-026. Compactability 54 4 As such, the deposits do “Influence of Nobake Core Sand Contamination on the Properties of not have a LOI value be- Base Sand System Parameters: 3% moisture, 6% seacoal, 8% Western Bentonite, 10 minute mulling time Green Molding Sands,” R.L. Naro, T. cause they will not burn in 1—From IMC Sand-Clay-Water Control Charts Zeh and J. Plummer, 1979 AFS a traditional LOI test. Based 2—Determined from Methylene Blue Titration Transactions, vol. 98, p. 39-46. MODERN CASTING / August 2004

37